Sample preparation for next-generation sequencing can involve fragmentation of genomic DNA or double-stranded cDNA (prepared from RNA) into smaller fragments, followed by addition of functional tag sequences (“tags”) to the strands of the fragments. Where a single-stranded sequence is tagged at both ends, the term “di-tagged” can be used. Such tags include priming sites for DNA polymerases for sequencing reactions, restriction sites, and domains for capture, amplification, detection, address, and transcription promoters. Previous methods for generating DNA fragment libraries required fragmenting the target DNA mechanically using a sonicator, nebulizer, or by a nuclease, and then joining (e.g., by ligation) the oligonucleotides containing the tags to the ends of the fragments.
A novel method for using transposons to rapidly achieve these steps was disclosed in US 2010/0120098 by Grunenwald, which is incorporated herein by reference, to generate fragments from any double-stranded DNA (e.g. genomic, amplicon, viral, phage, cDNA derived from RNA, etc.). Particularly useful transposon systems include the hyperactive Tn5 transposon system described in U.S. Pat. Nos. 5,965,443 and 6,437,109 by Reznikoff, and the Mu transposon system in U.S. Pat. No. 6,593,113 by Tenkanen, all of which are incorporated herein by reference. Reznikoff in particular described a 19-base transposase end sequence (SEQ ID NO:3) that is frequently referred to as “ME”. In some embodiments of the transposon method, polymerase chain reaction (PCR) is used as a downstream step for DNA amplification. This can raise concerns because of PCR's potential to over- or underrepresent the relative amounts of a given sequence, depending on its G+C composition, especially in regions of extreme G+C content where PCR bias can confound the annotation and analysis of the data.